Facebook Marketplace Scraper
Pricing
from $0.28 / 1,000 record scrapeds
Facebook Marketplace Scraper
Scrape Facebook Marketplace listings — titles, prices, descriptions, images, seller info, location, condition, and more. Supports search URLs, keyword search, city filtering, and detail page scraping.
Pricing
from $0.28 / 1,000 record scrapeds
Rating
0.0
(0)
Developer
kata Kuri
Maintained by CommunityActor stats
2
Bookmarked
67
Total users
6
Monthly active users
16 hours ago
Last modified
Categories
Features
Share
Reliable, no-login Facebook Marketplace scraper. Returns clean JSON with title, price, full description, photos, GPS coordinates, condition, category, and category-specific attributes (vehicle make/model/year/mileage, property bedrooms/bathrooms, and more).
Works without cookies for 90%+ of public listings. Most other scrapers go silent the moment Facebook rotates its SSR shape — this one walks the inlined JSON and merges every sibling target node, so you don't lose half the fields when FB changes things.
Why this scraper
| This actor | Typical alternative | |
|---|---|---|
| Anonymous mode | ✓ default — no cookies needed | Often "required", silently fails |
| Login wall handling | Content-aware: serves through soft overlays | Hard-fails on /login redirect |
| Description, GPS, photos | Extracted from inlined SSR JSON anonymously | Often null without cookies |
| SSR shape changes | Multi-source merge from every sibling target node | Returns partial / null when FB rotates |
| Honest completeness flag | _completeness: "basic" | "full" | Marks partial data as full |
| Cross-run dedup | Built-in monitoring mode | Add-on / missing |
| Incremental "since X" | untilDate filter | Missing |
| Strict keyword match | strictMatch: true removes broad-match noise | Missing |
| Auto same-country proxy | Picks US / GB / DE / FR / etc. from city | Manual config |
| Pricing (1K listings, all-in) | $0.40 search-only · $2.30 with full details — platform usage & proxy included | $0.50–$12.40 / 1K, often plus platform usage |
Quickstart
1. Search by keyword + city
{"searchQuery": "iphone 15","city": "nyc","maxResults": 50,"scrapeDetails": true,"proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] }}
2. Browse a category
{"city": "la","category": "furniture","maxResults": 100,"proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] }}
3. Drop in raw Marketplace URLs
{"searchUrls": ["https://www.facebook.com/marketplace/sanfrancisco/cars?minPrice=10000&maxPrice=30000","https://www.facebook.com/marketplace/seattle/electronics"],"scrapeDetails": true,"proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] }}
4. Incremental "what's new since yesterday"
{"searchQuery": "macbook pro","city": "austin","scrapeDetails": true,"monitoringMode": true,"monitoringStoreName": "fb-macbook-austin","untilDate": "2026-04-25","proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] }}
Schedule this hourly; you'll only ever pay for net-new listings.
Input configuration
Search
| Field | Type | Default | Description |
|---|---|---|---|
searchUrls | string[] | — | Paste any Marketplace URL: /marketplace/<city>/, /marketplace/<city>/<category>, /marketplace/<city>/search?query=.... Combinable with the search-builder fields. |
searchQuery | string | — | Keyword. Combined with city / category to build a URL. |
city | string | — | Facebook city slug (e.g. nyc, la, chicago, london, paris). Spellings like "New York" are mapped; other towns are geocoded to the nearest verified city; "lat,lng" also works. Facebook ignores unknown slugs and searches around the proxy IP — the run log warns when that happens. |
category | enum | — | vehicles, cars, phones, furniture, clothing, property-rentals, appliances, etc. |
sortBy | enum | best-match | newest, price-low, price-high, distance, best-match. |
condition | enum | — | new, used-like-new, used-good, used-fair. |
minPrice / maxPrice | int | — | Price filter (in the city's local currency). |
radiusKm | int | FB default | Search radius from city center, max 100 km. |
Scope and cost control
| Field | Type | Default | Description |
|---|---|---|---|
maxResults | int | 100 | Hard cap on listings returned (never exceeded). Changed 2026-11: leaving it empty (or null) now means 100, not unlimited. Set 0 for no cap: the search then stops by itself once Facebook stops showing new matching listings. |
scrapeDetails | bool | false | Visit each listing's detail page for description, GPS, photos, condition and the listing's Details (mileage, transmission…). Off by default (faster, cheaper card-only data); turn it on for full rows. |
maxConcurrency | int | 3 | 1–10. Anonymous runs need no browser: pages are fetched over plain HTTP with 4× this many parallel requests (max 20), each through its own proxy IP. With cookies or fullBrowserMode it is the number of parallel browser tabs — keep it low (2–3) there. |
Filters
| Field | Type | Default | Description |
|---|---|---|---|
strictMatch | bool | false | Drop any listing whose title doesn't contain every word of the query. Filters out FB's broad-match noise. |
untilDate | string | — | ISO date (2026-04-25). Skip listings posted before this date. Applied whenever publishedAt is known — mostly on detail pages, so use it with scrapeDetails: true. With sortBy: "newest" the search stops at the first older listing. |
Incremental scraping
| Field | Type | Default | Description |
|---|---|---|---|
monitoringMode | bool | false | Track listing IDs across runs in a named Key-Value store. Skip any IDs already delivered in earlier runs. IDs are saved during the run (not only at the end), so an aborted or timed-out run never re-charges you for listings you already received. |
monitoringStoreName | string | fb-marketplace-seen-<hash> | Name of the KV store. Empty = a store derived from the search URLs, so each scheduled search keeps its own watchlist. Set a name to share one watchlist or keep using an existing store (e.g. the old default fb-marketplace-seen; on the first run with the new default, IDs from that old store are imported automatically). |
Authentication (optional)
| Field | Type | Default | Description |
|---|---|---|---|
cookieState | string | — | Optional Facebook session cookies as JSON — Playwright storageState ({"cookies":[...]}) or the cookie array exported by EditThisCookie / Cookie-Editor. Anonymous works for public card and detail data; supply cookies only if you hit a login wall or need seller name/profile (FB hides those server-side anonymously). Stored as a secret. |
Infrastructure
| Field | Type | Default | Description |
|---|---|---|---|
proxyConfiguration | object | — | Standard Apify proxy config. Residential strongly recommended — datacenter IPs are blocked by Facebook. If you set a city and use residential without a countryCode, the actor auto-picks a same-country proxy. |
Output schema
Each row in the dataset:
{"listingId": "1279908470484354","title": "iPhone 12 Pro Max - Mint condition","price": 425,"currency": "USD","condition": "Used - Like New","category": "807311116002614","description": "Selling my iPhone 12 Pro Max, 256GB, mint condition...","city": "Austin","location": "Austin, TX","latitude": 30.267153,"longitude": -97.743057,"images": ["https://scontent.../photo1.jpg"],"sellerName": null,"sellerProfileUrl": null,"deliveryType": "IN_PERSON","listingUrl": "https://www.facebook.com/marketplace/item/1279908470484354/","publishedAt": "2026-04-30T23:00:11.000Z","scrapedAt": "2026-05-02T08:00:00.000Z","attributes": { "subtitles": ["256 GB", "Like new"] },"isSold": false,"isPending": false,"inputUrl": "https://www.facebook.com/marketplace/austin/search?query=iphone","_completeness": "full"}
Anonymous detail rows are read from Facebook's lightweight mobile listing page (about 5x less proxy traffic than the desktop page, which keeps the price low): they carry up to 5 photos — imageCount is the listing's total number of photos — and the listing's Details lines in attributes.details (vehicles also get mileage, mileageUnit, transmission, exteriorColor, interiorColor, owners). If that page can't be read, the desktop page is used and the fields below are filled.
Vehicle listings (desktop page, cookies or fullBrowserMode) additionally fill attributes.vehicleMake, vehicleModel, vehicleYear, mileage, mileageUnit, transmission, fuelType, exteriorColor, interiorColor, vehicleSellerType, vin.
Property listings additionally fill attributes.bedrooms, bathrooms, areaSize, areaUnit, propertyType, yearBuilt.
_completeness is "basic" when only card-level data was captured, "full" when detail-page data landed (a description, more than one photo, or category attributes). inputUrl is the search URL the listing came from. A listing whose detail page says it is no longer available is returned with its card data and isRemoved: true (basic, not detail-charged).
sellerName and sellerProfileUrl are null in anonymous mode — Facebook hides seller PII server-side without login.
Pricing
All-in pay-per-event pricing: platform usage and residential proxy are included — nothing else is billed to your account. Automatic Apify-plan discounts, no subscription, no minimums.
| Event | Free | Starter | Scale | Business |
|---|---|---|---|---|
| Actor start (per run, per GB of memory — default 1 GB) | $0.002/GB → $0.002 | $0.0018/GB | $0.0016/GB | $0.0014/GB |
| Record scraped (per 1,000) | $0.40 | $0.36 | $0.32 | $0.28 |
| Detail-enriched listing (per 1,000) | $1.90 | $1.00 ($1.80 from 30 Nov 2026) | $1.00 ($1.70 from 30 Nov 2026) | $1.00 ($1.60 from 30 Nov 2026) |
"Detail-enriched" is charged only with scrapeDetails: true and only when the detail page actually added a description, more than one photo, or category attributes. The run stops cleanly when your maximum total charge for the run is reached.
Typical run costs
| Mode | Free plan | Starter |
|---|---|---|
Search-only (1,000 listings, default scrapeDetails: false) | $0.40 | $0.36 |
Full detail (1,000 listings, scrapeDetails: true) | $2.30 | $1.36 ($2.16 from 30 Nov 2026) |
| Search-only (10,000 listings) | $4.00 | $3.60 |
Figures include everything: there is no separate platform-usage or proxy bill.
Cost vs alternatives (Free plan, 1K listings)
| Scraper | Search-only | Full detail |
|---|---|---|
| This actor | $0.40 | $2.30 |
| apify/facebook-marketplace-scraper | $6.20 | $12.40 |
| memo23/facebook-marketplace-scraper-ppe | $1.50 | $2.50+ |
| curious_coder/facebook-marketplace | $0.50 | $1.50 |
How it stays reliable
Three defensive layers, in priority order:
- Inlined SSR JSON — Facebook ships every page with multiple
<script type="application/json">blobs containing the full GraphQL responses. The actor scans every script, walks the JSON tree (up to 20 levels deep), and merges sibling target nodes so you don't lose half the fields when FB changes its shape. - Live
/api/graphqlinterceptor — when the page fires GraphQL calls (logged-in sessions or client-side re-fetches), those payloads are parsed too. Same parser, same field set. - Open Graph meta tags —
og:title,og:description,og:imageare always populated even if FB rotates its SSR shape.
Plus:
- No browser for anonymous runs — search pages and listing pages are fetched over plain HTTP with a real browser's TLS fingerprint and headers through the residential proxy. A browser (Playwright) is started only with cookies, with
fullBrowserMode, or to retry pages that plain HTTP could not read. - Consent-free listing pages — for EU/UK/Swiss searches, listing pages are fetched through US residential IPs, because Facebook shows logged-out visitors there a cookie-consent screen instead of the page.
- Per-context warm-up (browser mode) — visit
/marketplace/once before the real navigation so FB's bot detector sees a browser that browsed first. - Content-aware block detection — only fail on
/loginredirects when there's no marketplace data on the page; FB often serves the full data with a login overlay on top. - ID-matching guard — fields are only merged from items whose
idmatches the target listing, so the "Similar items" sidebar can't leak into your row. - Bounded scrolling — a search stops after a few scrolls without a new matching listing (Facebook keeps appending "results outside your search"), with a hard cap on scrolls and time per search, so runs never hang until timeout.
- Clear failures — if nothing could be scraped (login wall, blocked IP, no matching listings), the run fails with a message saying what to change instead of "succeeding" with an empty dataset. Failed URLs are listed in the key-value store record
FAILED_REQUESTS.
Common scenarios
Build a price-tracking feed
{"searchQuery": "iphone 15","city": "nyc","scrapeDetails": false,"monitoringMode": true,"monitoringStoreName": "fb-iphone15-nyc-prices"}
Schedule hourly. Each run only pushes listings you haven't seen — perfect for a (timestamp, listingId, price) time series.
Monitor a niche category
{"city": "boston","category": "musical-instruments","monitoringMode": true,"monitoringStoreName": "fb-instruments-boston","untilDate": "2026-04-01","scrapeDetails": true}
Strict comparison shopping
{"searchQuery": "macbook pro 14","city": "sanfrancisco","strictMatch": true,"scrapeDetails": true}
strictMatch drops any listing whose title doesn't contain both "macbook" and "pro" — no chargers, no docks, no accessories.
Used-car research
{"city": "dallas","category": "cars","minPrice": 5000,"maxPrice": 25000,"condition": "used-good","sortBy": "newest","scrapeDetails": true,"maxResults": 200}
Returns full vehicle attributes (vehicleMake, vehicleModel, vehicleYear, mileage, transmission, fuelType) for each listing.
FAQ
Issues / contact
Open an issue on this actor's page on Apify Store, or contact the maintainer through the Apify console.